Yubin Kuang
Co-Founder at Staer
Switzerland
Summary
🤩
Rockstar🎓
Top SchoolYubin Kuang is a founder and engineering leader with 11 years of experience building large-scale computer vision and spatial intelligence systems, currently leading a startup after senior roles at Meta Reality Labs and co-founding Mapillary. He combines PhD-level applied mathematics and practical ML engineering to ship production-ready perception models and infrastructure for world-scale 3D semantic mapping and on-device spatial perception. His hands-on open-source contributions to Mapillary projects (mapillary_tools and OpenSfM) show deep expertise in image processing, feature matching, and robust data handling for large imagery pipelines. Comfortable spanning research and product delivery, he has repeatedly translated academic rigor into scalable systems used in real-world mapping and AR platforms. Based in Switzerland, he brings a rare mix of entrepreneurship, core algorithmic fluency, and operational experience building perception foundations for consumer and mapping products.
11 years of coding experience
13 years of employment as a software developer
BSc. Earth Sciences, BSc. Earth Sciences at Nanjing UniversityDoctor of Philosophy - PhD Applied Mathematics, Doctor of Philosophy - PhD Applied Mathematics at Lund University
Chinese, Chinese, English
Github Skills (12)
image-data10
command-line-tool10
opencv10
metadata10
image-upload10
python10
feature-extraction10
3d-reconstruction10
algorithm9
algorithms9
cprogramming-language5
c-language5
Programming languages (5)
Github contributions (5)
mapillary/mapillary_tools
Sep 2014 - Oct 2018
Command line tools for processing and uploading Mapillary imagery
Role in this project:
Back-end Developer
Back-end DeveloperContributions:3 releases, 217 commits, 29 PRs in 4 years 1 month
Contributions summary:Yubin primarily contributed to the `mapillary_tools` repository by modifying and improving the Python-based command-line tools. Their work focused on refining image upload functionality, specifically removing a requirement and improving time-handling within the upload process. Additionally, they made enhancements to the processing of EXIF data within images, including fixing make/model fields and incorporating a shared library for common functionalities. These modifications demonstrate a focus on improving the core utilities for processing and uploading imagery.
clicommand-line-toolsimageryuploadingmapillary
mapillary/OpenSfM
Nov 2014 - Jun 2015
Open source Structure-from-Motion pipeline
Role in this project:Back-end Developer
Back-end DeveloperContributions:188 commits, 2 PRs, 69 pushes in 6 months
Contributions summary:Yubin primarily contributed to the `opensfm` repository by implementing and refactoring core feature extraction and matching components within the Structure-from-Motion pipeline. They added support for different feature types, including SURF and AKAZE, and optimized matching algorithms like brute-force matching. The commits also focused on core functionality, such as handling sensor databases and integrating focal length conversions, improving the robustness and flexibility of the image processing pipeline.
structure-from-motionpythonpipelinemotionintrinsics
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.